LTX-2

Official Python inference and LoRA trainer package for the LTX-2 audio–video generative model.

What is LTX-2?

Official Python inference and LoRA trainer package for the LTX-2 audio–video generative model.

Overview of LTX-2

What it is

LTX-2 is the first DiT-based audio-video foundation model that contains all core capabilities of modern video generation in one model: synchronized audio and video, high fidelity, multiple performance modes, production-ready outputs, API access, and open access. This repository provides the official Python inference and LoRA trainer package for the LTX-2 audio–video generative model.

Key features

  • Synchronized audio and video generation in a single model.
  • Multiple performance modes including a fast distilled pipeline and a production-quality DFR (Diffusion Fidelity Rendering) pipeline.
  • Supports text-to-video and image-to-video generation.
  • Includes a video VAE with a diffusion decoder for improved quality.
  • Offers LoRA training and inference capabilities.
  • Provides a spatial upscaler for higher resolution outputs.
  • Uses a Gemma 4 12B text encoder fine-tuned for LTX.

Use cases

  • Generating synchronized audio and video content from text prompts.
  • Production-quality video generation for media and entertainment.
  • Rapid prototyping with the fast distilled pipeline.
  • Fine-tuning the model with LoRA for custom styles or subjects.

Getting started

See the repository README for detailed installation and usage instructions. The quick start involves cloning the repository, installing dependencies with uv sync --extra natten, downloading model weights from Hugging Face, and running a generation command.

License and pricing

The code in this repository is open source under the Apache License 2.0. The LTX models are released under their own licenses: LTX-2.5 is under the LTX-2.x Community License, and LTX-2.3 is under the LTX-2 Community License. The model licenses apply whenever you use the models, including with this code.

Based on the project's public README and repository metadata.

Add this badge to your README

Featured on YunoDB Badge styles & help

Maintain this project? Show that it is featured on YunoDB.

Markdown[![Featured on YunoDB](https://yunodb.com/badge/lightricksltx-2.svg)](https://yunodb.com/tools/lightricksltx-2?utm_source=github&utm_medium=badge)
HTML<a href="https://yunodb.com/tools/lightricksltx-2?utm_source=github&utm_medium=badge"><img src="https://yunodb.com/badge/lightricksltx-2.svg" alt="Featured on YunoDB" height="20"></a>